Deepaware: A hybrid deep learning and context-aware heuristics-based model for atrial fibrillation detection

Research output: Contribution to journalJournal articlepeer-review

Standard

Deepaware : A hybrid deep learning and context-aware heuristics-based model for atrial fibrillation detection. / Kumar, Devender; Peimankar, Abdolrahman; Sharma, Kamal; Domínguez, Helena; Puthusserypady, Sadasivan; Bardram, Jakob E.

In: Computer Methods and Programs in Biomedicine, Vol. 221, 106899, 2022.

Research output: Contribution to journalJournal articlepeer-review

Harvard

Kumar, D, Peimankar, A, Sharma, K, Domínguez, H, Puthusserypady, S & Bardram, JE 2022, 'Deepaware: A hybrid deep learning and context-aware heuristics-based model for atrial fibrillation detection', Computer Methods and Programs in Biomedicine, vol. 221, 106899. https://doi.org/10.1016/j.cmpb.2022.106899

APA

Kumar, D., Peimankar, A., Sharma, K., Domínguez, H., Puthusserypady, S., & Bardram, J. E. (2022). Deepaware: A hybrid deep learning and context-aware heuristics-based model for atrial fibrillation detection. Computer Methods and Programs in Biomedicine, 221, [106899]. https://doi.org/10.1016/j.cmpb.2022.106899

Vancouver

Kumar D, Peimankar A, Sharma K, Domínguez H, Puthusserypady S, Bardram JE. Deepaware: A hybrid deep learning and context-aware heuristics-based model for atrial fibrillation detection. Computer Methods and Programs in Biomedicine. 2022;221. 106899. https://doi.org/10.1016/j.cmpb.2022.106899

Author

Kumar, Devender ; Peimankar, Abdolrahman ; Sharma, Kamal ; Domínguez, Helena ; Puthusserypady, Sadasivan ; Bardram, Jakob E. / Deepaware : A hybrid deep learning and context-aware heuristics-based model for atrial fibrillation detection. In: Computer Methods and Programs in Biomedicine. 2022 ; Vol. 221.

Bibtex

@article{20a9bc072d8e4089afd0438a2c022735,
title = "Deepaware: A hybrid deep learning and context-aware heuristics-based model for atrial fibrillation detection",
abstract = "Background: State-of-the-art automatic atrial fibrillation (AF) detection models trained on RR-interval (RRI) features generally produce high performance on standard benchmark electrocardiogram (ECG) AF datasets. These models, however, result in a significantly high false positive rates (FPRs) when applied on ECG data collected under free-living ambulatory conditions and in the presence of non-AF arrhythmias. Method: This paper proposes DeepAware, a novel hybrid model combining deep learning (DL) and context-aware heuristics (CAH), which reduces the FPR effectively and improves the AF detection performance on participant-operated ambulatory ECG from free-living conditions. It exploits the RRI and P-wave features, as well as the contextual features from the ambulatory ECG. Results: DeepAware is shown to be very generalizable and superior to the state-of-the-art models when applied on unseen benchmark ECG AF datasets. Most importantly, the model is able to detect AF efficiently when applied on participant-operated ambulatory ECG recordings from free-living conditions and has achieved a sensitivity (Se), specificity (Sp), and accuracy (Acc) of 97.94%, 98.39%, 98.06%, respectively. Results also demonstrate the effect of atrial activity analysis (via P-waves detection) and CAH in reducing the FPR over the RRI features-based AF detection model. Conclusions: The proposed DeepAware model can substantially reduce the physician's workload of manually reviewing the false positives (FPs) and facilitate long-term ambulatory monitoring for early detection of AF.",
keywords = "Arrhythmia, Atrial fibrillation, Context-awareness, Convolutional neural networks, Deep learning, Electrocardiogram (ECG), Health informatics, Long short-term memory (LSTM)",
author = "Devender Kumar and Abdolrahman Peimankar and Kamal Sharma and Helena Dom{\'i}nguez and Sadasivan Puthusserypady and Bardram, {Jakob E.}",
note = "Publisher Copyright: {\textcopyright} 2022 The Author(s)",
year = "2022",
doi = "10.1016/j.cmpb.2022.106899",
language = "English",
volume = "221",
journal = "Computer Methods and Programs in Biomedicine",
issn = "0169-2607",
publisher = "Elsevier Ireland Ltd",

}

RIS

TY - JOUR

T1 - Deepaware

T2 - A hybrid deep learning and context-aware heuristics-based model for atrial fibrillation detection

AU - Kumar, Devender

AU - Peimankar, Abdolrahman

AU - Sharma, Kamal

AU - Domínguez, Helena

AU - Puthusserypady, Sadasivan

AU - Bardram, Jakob E.

N1 - Publisher Copyright: © 2022 The Author(s)

PY - 2022

Y1 - 2022

N2 - Background: State-of-the-art automatic atrial fibrillation (AF) detection models trained on RR-interval (RRI) features generally produce high performance on standard benchmark electrocardiogram (ECG) AF datasets. These models, however, result in a significantly high false positive rates (FPRs) when applied on ECG data collected under free-living ambulatory conditions and in the presence of non-AF arrhythmias. Method: This paper proposes DeepAware, a novel hybrid model combining deep learning (DL) and context-aware heuristics (CAH), which reduces the FPR effectively and improves the AF detection performance on participant-operated ambulatory ECG from free-living conditions. It exploits the RRI and P-wave features, as well as the contextual features from the ambulatory ECG. Results: DeepAware is shown to be very generalizable and superior to the state-of-the-art models when applied on unseen benchmark ECG AF datasets. Most importantly, the model is able to detect AF efficiently when applied on participant-operated ambulatory ECG recordings from free-living conditions and has achieved a sensitivity (Se), specificity (Sp), and accuracy (Acc) of 97.94%, 98.39%, 98.06%, respectively. Results also demonstrate the effect of atrial activity analysis (via P-waves detection) and CAH in reducing the FPR over the RRI features-based AF detection model. Conclusions: The proposed DeepAware model can substantially reduce the physician's workload of manually reviewing the false positives (FPs) and facilitate long-term ambulatory monitoring for early detection of AF.

AB - Background: State-of-the-art automatic atrial fibrillation (AF) detection models trained on RR-interval (RRI) features generally produce high performance on standard benchmark electrocardiogram (ECG) AF datasets. These models, however, result in a significantly high false positive rates (FPRs) when applied on ECG data collected under free-living ambulatory conditions and in the presence of non-AF arrhythmias. Method: This paper proposes DeepAware, a novel hybrid model combining deep learning (DL) and context-aware heuristics (CAH), which reduces the FPR effectively and improves the AF detection performance on participant-operated ambulatory ECG from free-living conditions. It exploits the RRI and P-wave features, as well as the contextual features from the ambulatory ECG. Results: DeepAware is shown to be very generalizable and superior to the state-of-the-art models when applied on unseen benchmark ECG AF datasets. Most importantly, the model is able to detect AF efficiently when applied on participant-operated ambulatory ECG recordings from free-living conditions and has achieved a sensitivity (Se), specificity (Sp), and accuracy (Acc) of 97.94%, 98.39%, 98.06%, respectively. Results also demonstrate the effect of atrial activity analysis (via P-waves detection) and CAH in reducing the FPR over the RRI features-based AF detection model. Conclusions: The proposed DeepAware model can substantially reduce the physician's workload of manually reviewing the false positives (FPs) and facilitate long-term ambulatory monitoring for early detection of AF.

KW - Arrhythmia

KW - Atrial fibrillation

KW - Context-awareness

KW - Convolutional neural networks

KW - Deep learning

KW - Electrocardiogram (ECG)

KW - Health informatics

KW - Long short-term memory (LSTM)

UR - http://www.scopus.com/inward/record.url?scp=85131144412&partnerID=8YFLogxK

U2 - 10.1016/j.cmpb.2022.106899

DO - 10.1016/j.cmpb.2022.106899

M3 - Journal article

C2 - 35640394

AN - SCOPUS:85131144412

VL - 221

JO - Computer Methods and Programs in Biomedicine

JF - Computer Methods and Programs in Biomedicine

SN - 0169-2607

M1 - 106899

ER -

ID: 314072581